Theories of cognitive aging can be divided into two classes. Global theories of cognitive aging posit a small number of global effects of aging on information processing such as a generalized slowing (GS) of information processing. It is proposed that progress in distinguishing between these two classes of theories will be easier with the development and use of general models of information processing that allow one to examine the nature of age differences across a wide variety of tasks. These models allow researchers to detect general influences of aging on information processing and simultaneously provide a framework for interpreting local deviations from global trends. In addition, the application of general computation models helps determine the particular empirical results that will be maximally informative in understanding age differences in information processing. Random walk models are argued to be well suited to explore age differences in a variety of processing domains. These models provide the framework for interpreting age differences in performance and provide a unified framework for modeling age differences in both speed and accuracy measures. Previous applications of random walk models have shown that empirical results that have been viewed as supporting GS theories actually underdetermine the nature of age differences in processing. Based on the application of the random walk model, it is argued that analyses of response time (RT) distributions will be critical in constraining the development of more detailed models of cognitive aging. In this light, the proposed research will sample age differences in a variety of processing domains. A method for modeling the RT distributions in younger and older adults that involves fitting a theoretical function called an ex-Gaussian distribution to the empirical RT distributions will be used to quantify these age differences. Random walk models will be used to model age differences in performance and provide a framework for interpreting age differences in RT distributions. This program of research will provide information about global and local effects of aging, and will provide a more clear specification of the empirical slowing phenomena that is ubiquitous in cognitive aging research. Furthermore, this modeling approach will help to bridge the gap between cognitive aging models, and more traditional models of information processing.